Epilepsy affects approximately 2.5 million Americans, resulting in substantial disability. Because up to 30% of patients with epilepsy continue to have seizures despite appropriate treatment with antiepileptic medications, additional interventions to improve seizure control are needed. One approach to improving seizure control is to treat coexisting sleep disorders, such as obstructive sleep apnea. Obstructive sleep apnea (OSA) may exacerbate seizures via sleep fragmentation, sleep deprivation, or other pathophysiological processes that have not yet been determined. The investigators recently documented that OSA is common in epilepsy patients with seizures refractory to medical treatment. In addition, preliminary data in the form of retrospective case series by the investigators and others have suggested that treatment of OSA may improve seizure control. However, no prospective studies have been done to verify these findings. Proof that treating OSA is effective in reducing seizure frequency will require a multicenter Randomized Clinical Trial (RCT). This large RCT will test the hypothesis that treatment of OSA in patients with epilepsy refractory to medical treatment will reduce seizure frequency. In addition, the RCT will assess the impact of treating OSA on health-related quality of life and on daytime sleepiness, common concerns in epilepsy patients that are often attributable to antiepileptic medications or to frequent seizures rather than to a coexisting sleep disorder. The proposed aims of the Pilot Clinical Trial (PCT) are to determine critical information for the design of the RCT to allow for the testing of the above hypotheses in the RCT. In the PCT subjects 18 years and older with 4 or more seizures per month who meet survey criteria for OSA and other study criteria will be recruited at 3 different sites from epilepsy patients seen in clinical settings. A total of 60 subjects will be observed longitudinally through PSG confirmation and treatment of OSA and randomized to either therapeutic continuous positive airway pressure (CPAP) or sub-therapeutic (placebo or sham) CPAP in order to determine tolerability. Rates of adherence to therapeutic and sham CPAP and dropout rates due to antiepileptic drug changes during the treatment phase will be estimated. Specifically, the proposed PCT will: 1. Evaluate screening ranges on the Sleep Apnea scale of the Sleep Disorders Questionnaire (DA/SDQ), a survey instrument that is used to determine whether subjects are eligible for inclusion into the RCT. 2. Determine the necessity of performing two nights of PSG in patients with epilepsy. A second night of study increases the cost and may decrease recruitment in the RCT, but may be important to include given the night-to-night variability in the PSG and the potential for seizure occurrence during recordings. The working hypothesis is that one night of PSG will be sufficient for the RCT. 3. Determine rates of adherence to therapeutic and sham CPAP, dropout rates due to antiepileptic drug changes, and response rates will provide valuable data for planning the RCT. 4. Develop quality control measures to ensure accurate and consistent data collection among sites in the RCT, including aspects related to remote data entry and standardization of performance and interpretation of PSG studies across sites.